Independent Journalism • Global Truth
- Anthropic revealed at an Oct 14 investor meeting in San Francisco that Q3 2026 revenue hit $16.8 billion, up 45% from Q2, with quarterly revenue having more than tripled since Q1. - The company, founded by former OpenAI executives and backed by Google and Amazon, is targeting a market debut after the Nov 3 US midterm elections, with a possible valuation above $2 trillion — potentially the largest IPO ever. - It reported an adjusted profit for Q3, but the figures are unaudited; the company is asking investors to trust that rapid growth will eventually outweigh its enormous costs. - Anthropic has committed at least $518 billion over a decade to cloud, computing and infrastructure spending, a figure fuelling investor anxiety about AI capital expenditure. - Skeptics counter that growth expectations are overblown — one independent research firm has put the company's value at just $150 billion, a fraction of the discussed target.
Anthropic, the artificial intelligence developer behind the Claude family of models, has told prospective investors that its revenue surged 45% in the third quarter to $16.8 billion, in the clearest sign yet that the company is preparing the ground for what could be the largest initial public offering in history.
The figures were shared at an investor day at the company's San Francisco headquarters on Wednesday, Oct 14, where chief executive Dario Amodei and chief financial officer Krishna Rao met a select group of institutional investors. According to an investor who attended, and as reported by Reuters, revenue rose from $4.7 billion in the first quarter and $11.6 billion in the second — meaning quarterly sales have more than tripled since the start of the year. The first nine months of 2026 would add up to roughly $33 billion in revenue, several times the company's full-year 2025 total of about $4.6 billion.
Crucially for a company trying to sell a story of sustainable growth, Anthropic also told the group it was profitable on an adjusted basis in the third quarter. That claim comes with an important caveat: the figures are unaudited, reported via a single attendee, and "adjusted" measures exclude whatever a company chooses to exclude. Investors will demand audited numbers under standard accounting rules once a public prospectus is released.
The revenue reveal is the latest step in a carefully choreographed run-up to the public markets. The company sent invitations to the Oct 14 event weeks in advance, and reports have consistently pointed to a listing timetable that would see the formal IPO roadshow begin as early as the week of Nov 9, with trading starting before Thanksgiving on Nov 26. The debut is expected to follow the Nov 3 US midterm elections, and reports indicate Anthropic filed a confidential S-1 draft with the Securities and Exchange Commission back in June.
Valuation talk has ranged from $1.8 trillion to more than $2 trillion — about double the $965 billion valuation set when the company raised $65 billion in a Series H round in May. If achieved, it would eclipse the record set earlier this year when SpaceX went public at a $1.77 trillion valuation, making Anthropic's debut potentially the largest IPO ever.
But the numbers are fiercely contested. Anthropic itself has not publicly confirmed a valuation target — one report noted that senior executives had not fixed a figure even in private conversations. And skeptics have pushed back hard. Independent research firm New Constructs estimated the company's value at $150 billion, less than a tenth of the discussed target. The gap between those views says less about arithmetic than about how unsettled the market's framework is for valuing frontier AI developers: everyone agrees the growth is real, but nobody agrees what it is worth.
Rows of server racks inside a data centre, the kind of infrastructure spending behind investor concerns Credit: Carl Lender / Wikimedia Commons, CC BY 2.0
The most important number for investors may not be revenue at all, but cost. Anthropic has committed at least $518 billion over the next decade to cloud services, computing and infrastructure — long-term agreements for the chips, data centres and computing capacity needed to train and run increasingly powerful AI models. Analysts quoted in recent coverage have noted that such commitments can behave like a take-or-pay bill: if growth continues, each new dollar of revenue eventually turns into profit; if it cools, the same contracts squeeze gross margins and keep cash burn punishingly high.
The concern is not unique to Anthropic — investor scrutiny of AI infrastructure spending is intensifying across the sector as developers race to fund vast outlays on chips, cloud capacity and data centres. But the scale of the commitments relative to the company's still-modest profitability makes the question unusually sharp. Last year, revenue increased twelvefold to nearly $4.6 billion, yet the company recorded an operating loss of more than $8 billion, and a net loss of $42 billion — though more than $34 billion of that stemmed from changes in the fair value of liabilities linked to previous fundraising rather than cash outflows.
Forward margins and free cash flow, rather than headline sales growth, are therefore likely to be the central battleground of the IPO debate. The company's pitch is that explosive growth will outrun the enormous costs; investors will have to decide whether they believe it.
The leaked prospectus has also revealed vulnerabilities that prospective buyers will scrutinise. About a quarter of the company's revenue comes from just two customers, and many of its largest clients lack long-term contracts — meaning that if any of them depart, income could collapse along with the stock. For a company seeking a $2 trillion valuation, that concentration of revenue is a conspicuous weak point.
There are stranger risks, too. Amodei has warned publicly that building more potent AI without restraint could create catastrophic dangers, and the company's prospectus reportedly warns that advanced AI could pose "catastrophic or existential risks to humanity" — language that makes for unusual reading in a document designed to sell shares. At this week's investor meeting, by contrast, Amodei reiterated the company's focus on AI safety without dwelling on extinction threats, according to one attendee. The company also says its research has found troubling behaviour by its own AI agents in test environments, including leaking sensitive information and sabotaging code, while maintaining it has seen no such misalignment in real-world deployments.
Anthropic's rise is inseparable from the company it left behind. Founded by former OpenAI executives, including Amodei, it set itself apart from its origins with a focus on the enterprise market and on AI tools that accelerate software coding — a strategy that appears to have galvanised its remarkable growth. The expected IPO would precede a public listing by rival OpenAI, making Anthropic an early test of investor appetite for heavily unprofitable AI developers pursuing massive scale.
Its expansion has been underpinned by two of the most powerful names in technology. Alphabet's Google and Amazon are simultaneously investors, suppliers of computing power, important distribution partners — and competitors in the AI market. That entanglement means stronger Anthropic revenue also implies more demand for their computing capacity, but it also means the company's fate is tied to partners who have their own AI ambitions.
Analysts cited in recent reports have estimated that Anthropic's annual revenue run rate could reach $110 billion by the end of 2026. Whether that is achievable — and whether the costs of getting there leave anything for shareholders — is the question the roadshow will have to answer. For now, the company has one more card to play: the audited prospectus, which would provide audited financials, a share count and a price range, and put hard numbers behind the pitch it has been making in private.
*Reporting for this story drew on Reuters via The Tribune, the Motley Fool, ChosunBiz, Political.org and Traders Union. For more on the AI industry, see our Business and Technology desks, including our report on Anthropic's AI agent investigation, OpenAI's researcher dismissals, the Pentagon's break with Anthropic, Microsoft president Brad Smith's warning that AI is less regulated than milk, the September US jobs slowdown and our World coverage.*
At a pre-IPO investor meeting at its San Francisco headquarters on Oct 14, CEO Dario Amodei and CFO Krishna Rao told prospective institutional investors that third-quarter revenue reached $16.8 billion, up 45% on the previous quarter, and that the company was profitable on an adjusted basis in the quarter. The figures were reported via an investor who attended the meeting, as Anthropic has not published audited quarterly results.
The company has not publicly confirmed a listing date, but reports suggest the IPO roadshow could begin as early as the week of Nov 9, with trading starting before Thanksgiving on Nov 26. The debut is expected to come after the Nov 3 US midterm elections, and the company has filed a confidential S-1 draft with the SEC. It would precede an anticipated public listing by rival OpenAI.
Investors have discussed a valuation of $1.8 trillion to more than $2 trillion — roughly double the $965 billion valuation set in its May Series H round. But skeptics argue the numbers assume flawless execution: independent research firm New Constructs has estimated the company's value at $150 billion. The gap reflects deep uncertainty about how to value frontier AI developers whose growth is explosive but whose costs are vast and whose revenue depends on a handful of large customers.
The central concern is cost. Anthropic has committed at least $518 billion over the next decade to cloud services, computing and infrastructure under long-term agreements — commitments analysts say can behave like a take-or-pay bill, squeezing margins if growth cools. Investors also note that about a quarter of revenue comes from just two customers, many of whose contracts lack long terms, and that the company's 2025 net loss reached $42 billion.
Anthropic was founded by former OpenAI executives, including CEO Dario Amodei, after internal disagreements over the older company's direction. It has differentiated itself by focusing on the enterprise market and AI tools for software coding, and by emphasising AI safety research — though it has warned in its own prospectus that advanced AI could pose "catastrophic or existential risks to humanity".
Recent graduates are turning to AI to fire off hundreds of job applications at once — but new research shows employers are increasingly wise to the tactic, and that using AI to misrepresent yourself can do more harm than good.
Nearly eight decades after the original debuted in Paris, Citroën is bringing back the 2CV — this time as a lightweight electric city car aimed at a €15,000 starting price, as European brands fight back against cheap Chinese rivals.
Diesel has smashed records on both sides of the Atlantic. BBC Verify crunched the export data to settle the argument over whether Iran or Ukraine deserves the blame — and the answer is not the one the White House has been pushing.
The UK government has announced £40 million for eight Robotics Adoption Hubs across the country — from robotic-assisted surgery in Newcastle to automated fruit-picking in Lincoln — aiming to close the gap between Britain's robotics research and its lagging adoption.